The invention relates to the production of a metal, such as aluminum, by electrolysis of a compound of the metal and in particular to the production of aluminum by electrolysis of a compound thereof, such as chloroaluminate or fluoroaluminate. More particularly, the invention relates to graphite containing composite cathode electrodes or electrode members containing a refractory hard metal, such as TiB.sub.2 used in such electrolytic cells, and to the manufacture of such electrode members and to selective use thereof in electrolytic cells.
Generally speaking, aluminum is produced by electrolysis of aluminum compounds, such as aluminum oxides or salts or other compounds, in a molten salt bath. Such usually concerns situating the molten salt electrolyte between an anode and cathode and passing current through the gap between the anode and cathode. One of the more prominent of such baths is the fluoroaluminate bath (AlF.sub.3 --NaF--CaF.sub.2) used in the well-known Hall cell. Another type of bath is the chloroaluminate type (AlCl.sub.3 --NaCl--LiCl--KCl--MgCl.sub.2) used in other cells such as is described in U.S. Pat. No. 3,755,099. In electrolytic cells for the production of aluminum, it is common for the anode to be vertically spaced from the cathode such that the current passes in a generally vertical direction through the bath. The anode can be a baked carbon block and the cathode, as seen by the salt bath, is liquid aluminum. Current passes from the anode through the salt bath to the liquid aluminum cathode and thence to the supporting media beneath the liquid aluminum (typically the bottom of the cell). The cells may be monopolar, such as depicted in U.S. Pat. Nos. 3,400,061 and 4,071,420, or they can be bipolar, such as depicted in U.S. Pat. No. 3,755,099, all of which are incorporated herein by reference.
One problem in the operation of such electrolytic cells in producing aluminum is the desire to increase the power efficiency in operating the cell. This could be accomplished by decreasing the distance beween the anode and the cathode, thus lessening the electrical resistace within the cell. However, such disturbs the liquid aluminum cathode and causes it to vary in thickness, thus limiting the anode-cathode spacing to that which safely accommodates the highest waves in the liquid aluminum cathode without short-circuiting.
It is recognized in the art that the use of refractory hard metal, such as titanium diboride, as a cathode surface in such cells offers significant advantages since the TiB.sub.2 surface is readily wettable by liquid aluminum. This then permits the cathode surface to be drained of liquid aluminum and elminates the problems caused by electromagnetic disturbance of a deeper or thicker aluminum pool and enables achieving reduced anode-cathode distances and improved power efficiencies. However, achieving a dependable titanium diboride cathode surface has often been difficult and quite expensive. Not only is titanium diboride very expensive, but it suffers from sensitivity to thermal stress encountered in heating a cell to start production. Composites of carbon and TiB.sub.2 including TiB.sub.2 -coated carbon have been tried but results to date have not been entirely satisfactory.